madebyollin

Variational Autoencoders Will Never Work

So you want to generate images with neural networks. You're in luck! VAEs are here to save the day. They're simple to implement, they generate images in one inference step (unlike those awful slow autoregressive models) and (most importantly) VAEs are 🚀🎉🎂🥳 theoretically grounded 🚀🎉🎂🥳 (unlike those scary GANs - don't look at the GANs)!

The idea

The idea of VAE is so simple, even an AI chatbot could explain it:

1. Your goal is to train a "decoder" neural network that consumes blobs of random noise from a fixed distribution (like torch.randn(1024)), interprets that noise as decisions about what to generate, and produces corresponding real-looking images. You want to train this network with nice simple image-space MSE loss against your dataset of real images.

looopTools

    inline std::vector<uint8_t> read_vector_from_disk(std::string file_path)
    {
        std::ifstream instream(file_path, std::ios::in | std::ios::binary);
        std::vector<uint8_t> data((std::istreambuf_iterator<char>(instream)), std::istreambuf_iterator<char>());
        return data;
}

fakuivan

Dishonored2.exe+3BFC7F0 | aas_subdivisionSize = 64 | Dishonored2.exe+3BFC818 -> 64 
the size of subdivisions to use for debug drawing 
 
Dishonored2.exe+340CC20 | achievements_Verbose = 0 | Dishonored2.exe+340CC48 -> 0 
debug spam for achievements 
 
Dishonored2.exe+3BFA570 | ai_debugCam = 0 | Dishonored2.exe+3BFA598 -> 0 
enable debug camera 
 
Dishonored2.exe+3BFA470 | ai_debugScript = -1 | Dishonored2.exe+3BFA498 -> 4294967295 

evanthebouncy

# An interactive game where the computer can learn to recognize
# more animals!

# Not sure if this is best practice for clarity, but i kept the
# data structure more like data rather than objects with methods
# so the recursion looks cleaner.

# an animal, like a dog or a cat
class Animal:
  def __init__(self, animal):